Pith. sign in

REVIEW 3 major objections 5 minor 41 references

Decomposability-Guaranteed Cooperative Coevolution for Large-Scale Itinerary Planning

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proves a sufficient condition under which large-scale itinerary planning can be decomposed city by city, then builds a cooperative coevolution algorithm that reports higher hypervolume than four comparison methods on all 18…

desk verdict The decomposability guarantee only proves existence of some cluster-once route, yet DGCC fixes one city order and never reorders, so the theoretical and algorithmic claims don't match; the empirical work is solid but the theory needs major revision. read the letter →

arxiv 2506.06121 v2 pith:D6H3Q22K submitted 2025-06-06 cs.AI

Ziyu Zhang , Peilan Xu , Yuetong Sun , Yuhui Shi , Wenjian Luo This is my paper
classification cs.AI
keywords itineraryplanningcooperativecoevolutionweakdecomposabilitylarge-scalemulti-objectiveoptimizationtravelingsalesmanproblemorienteeringNSGA-IIhypervolume
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Large-scale itinerary planning—choosing a route across many cities that maximizes collected point-of-interest scores while minimizing travel time and cost—can be decomposed city by city without giving up optimality, provided intercity travel links are heavy compared with links inside each city. The paper proves a sufficient condition (Theorem 1) under which some optimal route visits each city exactly once as a contiguous block, calling this weak decomposability. It then builds DGCC, a cooperative coevolution algorithm that optimizes each city as a separate component, dynamically shifts travel days between cities using hypervolume of normalized fitness, and distributes computation by an optimization-potential measure. In experiments on 18 real-world itinerary cases from China, France, and Germany, DGCC reports higher hypervolume than four comparison algorithms on every instance, with the advantage increasing as the number of cities grows.

What carries the argument

The load-bearing object is the inequality in Theorem 1: the sum, over all cities, of the maximum edge weight inside each city is at most the minimum edge weight between any two cities. When it holds, the exchange argument in the theorem shows any route can be reordered so that each city is visited in one contiguous block without increasing the objective, making the problem weakly decomposable. On that foundation, DGCC uses a per-day encoding with placeholder zeros, a dynamic decomposition that moves encoding units from the city with the lowest normalized-fitness hypervolume to the highest, and a computational resource allocation where each city's optimization potential is the product of its hypervolume improvement and its POI count, with stagnant cities receiving only basic evaluations.

What would settle it

For each of the 18 test cases, compute every intra-city edge weight and every inter-city edge weight from the travel data and test whether the sum of the within-city maxima is at most the minimum between-city edge; any violation breaks the weak-decomposability guarantee. Then run an unrestricted multi-objective optimizer without the city-block constraint and compare its Pareto front with DGCC's; a strictly better unrestricted front would show that the fixed city-order encoding excludes optimal solutions.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that multi-city itinerary planning is weakly decomposable: under the condition that the sum of the largest travel edges inside the cities is no larger than the smallest edge between cities, there always exists an optimal route that visits each city exactly once, so the route can be rearranged into consecutive per-city blocks. Strict decomposability, in which each city's optimal subpath is independent of the others, is shown to be generically impossible. The paper then treats this guarantee as a license to decompose by city and proposes DGCC, whose components are per-city subpopulations with a travel-day encoding, whose dynamic decomposition transfers encoding units from low-contribution to high-contribution cities, and whose resource allocation uses hypervolume improvement combined with city POI count. The reported result is that DGCC significantly outperforms CCIP, LSNSGA-II, MOPSO, and AONSGA-II on all 18 test cases under the Wilcoxon test, with performance advantages that grow with problem scale.

Load-bearing premise

The load-bearing premise is the unverified assertion that every test case satisfies the Theorem 1 inequality—the largest within-city travel links are collectively no bigger than the smallest between-city link—and the extra assumption that the fixed city order used by DGCC contains an optimal route.

Editorial extensions

If this is right

  • When the Theorem 1 inequality holds, searching only over routes that visit each city as one contiguous block does not discard all optimal solutions.
  • Giving more travel days to cities with richer POI sets through dynamic decomposition directly targets the imbalance that the ablation study shows is important.
  • The reported superiority over CCIP, LSNSGA-II, MOPSO, and AONSGA-II on every one of the 18 cases suggests a generic advantage for decomposable multi-city itineraries.
  • The performance gap growing with the number of cities supports the claim that the curse of dimensionality is mitigated by city-level decomposition.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The assertion that all 18 test cases satisfy the Theorem 1 inequality is stated but not demonstrated in the paper; checking it against the travel-time and travel-cost matrices would settle whether the decomposability guarantee actually holds for the experiments.
  • The theorem guarantees that a contiguous-block route exists, not that the particular city order fixed by DGCC is optimal, so the fixed order is an extra assumption beyond the theorem.
  • The same condition-and-exchange argument should transfer to other clustered routing problems, such as clustered TSP with profits or multi-zone delivery routing, whenever inter-zone travel dominates intra-zone travel.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. This paper addresses large-scale itinerary planning as a multi-objective TSP variant in which POIs are clustered by city. It introduces a formal notion of weak decomposability, proves a sufficient graph condition (Theorem 1) under which an optimal path visits each cluster exactly once, and proposes DGCC, a cooperative coevolutionary algorithm with dynamic component-size adjustment and computational-resource allocation. Experiments are run on 18 real-world test cases built from POIs in Chinese, French, and German cities, and the authors report statistically significant hypervolume improvements over four baseline algorithms, with larger gains for more cities.

Significance. If the weak-decomposability guarantee were correctly established and actually used, the paper would be a useful bridge between problem structure and cooperative coevolution for itinerary planning. The paper has concrete strengths: real-world data, 50-run averages with Wilcoxon and Friedman tests, ablations of the three main algorithmic components, and parameter studies. The central theoretical-to-algorithmic link, however, is not established: Theorem 1 is not proved as written, the assertion that all test cases satisfy condition (2) is unverified, and the algorithm optimizes over a fixed city order that the theorem does not justify. These are load-bearing gaps in the 'decomposability-guaranteed' claim.

major comments (3)
  1. [Section III.A, Theorem 1] The base-case proof of Theorem 1 is not valid as written. For the paths x=(v(1)_1, v(2)_1, v(1)_2, v(2)_2) and x'=(v(1)_2, v(1)_1, v(2)_1, v(2)_2), the true difference f(x')-f(x) contains the term -w(v(2)_{1,e}, v(1)_{2,s}) (the edge between the second and third segments of x) and does not contain the term -w(v(1)_{1,e}, v(2)_{1,s}), because that edge is present in both paths and cancels. The displayed formula therefore does not equal the actual difference, and the claimed bound w_max(V_1)+w_max(V_2)-2w_min(V_1,V_2) is not derived. The induction step is a sketch that never accounts for all edges affected by the 'extract' and 'reinsert' operations. Theorem 1 is consequently not established in the manuscript.
  2. [Section III.A, final paragraph] The statement 'In this paper, the test cases satisfy the conditions of Theorem 1' is unsupported. No computation of w_max(V_i) or min_{j≠k} w_min(V_j,V_k) is reported for any of the 18 test cases, so the reader cannot verify the load-bearing premise that the experimental instances are weakly decomposable. Moreover, the experimental objectives in Eq. (7) differ from Eq. (1): they include vertex costs, a reciprocal POI score, and a k-dependent balancing factor ω in Eq. (8), while Theorem 1's condition is stated only for scalar edge weights. The paper should either report the verification for each case or explicitly extend the theorem to the actual objective vector.
  3. [Sections III.B, III.E, Algorithm 2] The weak-decomposability guarantee does not justify the fixed city order used by DGCC. Theorem 1 asserts the existence of an optimal path that visits each cluster exactly once, but it says nothing about which cluster order is optimal. Algorithm 2 line 3 fixes an 'optimal visiting sequence' at initialization, and the dynamic decomposition in Algorithm 1 only transfers encoding units without changing that order; the assembly step then concatenates component blocks in this fixed order. The paper provides no proof, and no dataset-specific argument, that this order contains a Pareto-optimal solution for any test case. The search space actually optimized is therefore a strict subset of the path space covered by Theorem 1, and the reported HV gains could be solving an order-restricted problem rather than the problem defined in Section II.A.
minor comments (5)
  1. [Section III.A, Lemma 1] The statement and proof of Lemma 1 contain an incomplete clause ('since no two path segments x^(i), we must have ...') and the condition on objective values is not stated precisely; the proof should say clearly that the argument assumes any two distinct path segments differ in objective value.
  2. [Algorithm 1, lines 11-18] The loop 'for i ← i_max to i_min' and the accompanying text assume i_max < i_min, but no justification is given; if i_max > i_min, the transfer direction is undefined. The pseudocode should specify the iteration direction for both cases.
  3. [Section IV.E, Table IV] The qualitative case study is inconsistent with the test-set definition: Table I assigns case 1 to Wuxi and Yangzhou, yet Section IV.E describes a four-day itinerary to Nanjing and Wuxi, and many listed POIs (e.g., Wangshi Garden, Zhuozheng Garden, Hanshan Temple, Guanqian Street) are in Suzhou rather than the stated cities. This inconsistency should be corrected and the underlying city labels checked.
  4. [Table III] The ablation results are reported without significance tests; 'outperforms' is inferred from mean HV only. Please add statistical comparisons or state explicitly that the differences are numerical.
  5. [Throughout] There are several presentation issues: 'German' is used as an adjective for Germany in Table I; the notation f(x^(i)) in Eq. (3) is used before being defined; and the meaning of the summary row '++/- 18/0/0' should be explained more explicitly in the text.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the weak-decomposability claim is proven in-paper and experimental results rest on external benchmarks; the flagged gaps (unverified condition (2) and the fixed component order) are correctness concerns, not reductions of outputs to inputs.

full rationale

The derivation chain is self-contained. The central theoretical claim (weak decomposability, Definition 2 / Theorem 1) is proven inside the paper by induction on the number of clusters; it is not imported from prior work or from the same-group CCIP paper [17]. The empirical claims (DGCC superior on 18 test cases) rest on comparisons with external algorithms — MOPSO, LSNSGA-II, AONSGA-II — plus the same-group CCIP as a baseline; the ablation study isolates each mechanism. Reference [17] (Z. Zhang, P. Xu, Z. Wang, W. Luo) shares authors with the present paper, but it is used only as related-work anchor and as an experimental baseline; it supplies no load-bearing premise, so it does not make the comparison circular. Two gaps were flagged but neither is circularity. (1) Section III.A asserts 'the test cases satisfy the conditions of Theorem 1' without verifying Eq. (2) against the dataset — a soundness/verifiability gap, since Eq. (2) is a sufficient condition and the real-world data may not satisfy it. (2) Theorem 1 guarantees only that some contiguous-cluster optimal path exists, whereas DGCC fixes the component visiting sequence at initialization (Algorithm 2, line 3: 'Initialize the optimal visiting sequence of components') and never reorders it, so the search space actually optimized is a subset of the weakly decomposable path space; this is an unproven assumption, not a circular step. Neither the hypervolume-based dynamic decomposition (Algorithm 1), the optimization-potential resource allocation (Eq. 4), nor the reported HV values is defined in terms of the paper's conclusions, and no fitted parameter is renamed as a prediction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on a strong, unverified data condition (Theorem 1) and on an unjustified fixed city order. The algorithm's hyperparameters are listed but some values (L, delta) are not fully reported. No new physical or conceptual entities are introduced beyond the weak-decomposability definition.

free parameters (5)
  • alpha (control parameter in objective) = 0.8
    Chosen by hand in Eq. (8) to balance the number of POIs in itineraries; not fitted to data but affects all objectives and results.
  • theta (scaling factor for Fe) = 10000
    Chosen by hand to adjust the range of the experience objective; listed in Section IV.A.
  • delta (balancing coefficient offset) = small positive constant, exact value not stated
    Introduced in Eq. (5) to prevent B^k from becoming zero; value is unspecified.
  • stagnation threshold = 5e-5
    Empirical threshold defining a stagnant component in Section III.C; no justification.
  • L (structure adjustment period) = not reported in main experiments
    Section IV.D analyzes L on test cases and finds peaks at L=6 to 11, but the value used in Table II is not stated, making the main results hard to reproduce.
assumptions (4)
  • domain assumption Each objective function in the test instances has the additive form of Eq. (1), i.e., a weighted sum of vertex and edge weights.
    Stated in Section III.A: 'each objective function follows the form of Equation (1)'. This is needed for the rearrangement argument of Theorem 1.
  • ad hoc to paper All test cases satisfy the sufficient condition of Theorem 1: sum of maximum intra-cluster edge weights is no larger than the minimum inter-cluster edge weight.
    Asserted without evidence in Section III.A; this licenses the weak decomposability guarantee and the contiguous-city search space.
  • ad hoc to paper The predefined city visiting order (Algorithm 2, step 3) contains an optimal or near-optimal ordering, or the algorithm can reorder the cities.
    The weak decomposability theorem only guarantees existence of some contiguous-cluster optimal path, not that a fixed order is optimal. The paper provides no mechanism to change the order.
  • standard math The graph is undirected and connected, with symmetric edge weights.
    Used in Lemma 1 and Theorem 1 proof; the underlying city network is modeled as undirected.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Decomposability-Guaranteed Cooperative Coevolution for Large-Scale Itinerary Planning." pith.science (2026). https://pith.science/paper/D6H3Q22K

@misc{pith2026250606121,
  author       = {Pith},
  title        = {Pith review of: Decomposability-Guaranteed Cooperative Coevolution for Large-Scale Itinerary Planning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D6H3Q22K}},
  note         = {Machine review of arXiv:2506.06121}
}
read the original abstract

Large-scale itinerary planning is a variant of the traveling salesman problem, aiming to determine an optimal path that maximizes the collected points of interest (POIs) scores while minimizing travel time and cost, subject to travel duration constraints. This paper analyzes the decomposability of large-scale itinerary planning, proving that strict decomposability is difficult to satisfy, and introduces a weak decomposability definition based on a necessary condition, deriving the corresponding graph structures that fulfill this property. With decomposability guaranteed, we propose a novel multi-objective cooperative coevolutionary algorithm for large-scale itinerary planning, addressing the challenges of component imbalance and interactions. Specifically, we design a dynamic decomposition strategy based on the normalized fitness within each component, define optimization potential considering component scale and contribution, and develop a computational resource allocation strategy. Finally, we evaluate the proposed algorithm on a set of real-world datasets. Comparative experiments with state-of-the-art multi-objective itinerary planning algorithms demonstrate the superiority of our approach, with performance advantages increasing as the problem scale grows.

Figures

Figures reproduced from arXiv: 2506.06121 by the authors.

Figure 1
Figure 1. The comparative schematic of the generalized TSP, clustered TSP, and large-scale itinerary planning. Orange backgrounds [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The schematic diagram of problem decomposition [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. The variation curve of normalized HV with respect [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The variation curve of normalized HV with respect [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

41 extracted references · 40 canonical work pages

  1. [1]

    Orienteering problem: A survey of recent variants, solution approaches and applications,

    A. Gunawan, H. C. Lau, and P. Vansteenwegen, “Orienteering problem: A survey of recent variants, solution approaches and applications,” European Journal of Operational Research, vol. 255, no. 2, pp. 315–332, 2016

    work page 2016

  2. [2]

    The orien- teering problem: A survey,

    P. Vansteenwegen, W. Souffriau, and D. Van Oudheusden, “The orien- teering problem: A survey,”European Journal of Operational Research, vol. 209, no. 1, pp. 1–10, 2011

    work page 2011

  3. [3]

    A development of travel itinerary planning application using traveling salesman problem and k- means clustering approach,

    S. Rani, K. N. Kholidah, and S. N. Huda, “A development of travel itinerary planning application using traveling salesman problem and k- means clustering approach,” inProceedings of the 2018 7th International Conference on Software and Computer Applications, 2018, pp. 327–331. 12 TABLE IV: Itineraries for DGCC, CCIP, and LSNSGA-II Algorithms Day City Poin...

    work page 2018

  4. [4]

    Automatic itinerary planning using triple-agent deep reinforcement learning,

    B.-H. Chen, J. Han, S. Chen, J.-L. Yin, and Z. Chen, “Automatic itinerary planning using triple-agent deep reinforcement learning,”IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 10, pp. 18 864–18 875, 2022

    work page 2022

  5. [5]

    Anytime planning of optimal schedules for a mobile sensing robot,

    J. Yu, J. Aslam, S. Karaman, and D. Rus, “Anytime planning of optimal schedules for a mobile sensing robot,” in2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2015, pp. 5279–5286

    work page 2015

  6. [6]

    A multi-user decision support system for online city bus tour planning,

    S. A. Bagloee, M. Tavana, D. Di Caprio, M. Asadi, and M. Heshmati, “A multi-user decision support system for online city bus tour planning,” Journal of Modern Transportation, vol. 25, pp. 59–73, 2017

    work page 2017

  7. [7]

    Research on travel route planning optimization based on large language model,

    B. Li, K. Zhang, Y . Sun, and J. Zou, “Research on travel route planning optimization based on large language model,” in2024 6th International Conference on Data-driven Optimization of Complex Systems (DOCS). IEEE, 2024, pp. 352–357

    work page 2024

  8. [8]

    An integrated recommender system for multi-day tourist itinerary,

    F. Ghobadi, A. Divsalar, H. Jandaghi, and R. B. Nozari, “An integrated recommender system for multi-day tourist itinerary,”Applied Soft Com- puting, vol. 149, p. 110942, 2023

    work page 2023

Show all 41 references
  1. [9]

    Recommendation of tourist itineraries with depen- dence on transport time,

    C. Gonz ´alez-Navasa, J. A. Moreno-P ´erez, J. Brito-Santana, and H. Alonso-Afonso, “Recommendation of tourist itineraries with depen- dence on transport time,”Transportation Research Procedia, vol. 71, pp. 77–84, 2023

    2023

  2. [10]

    Samap: An user-oriented adaptive system for planning tourist visits,

    L. Castillo, E. Armengol, E. Onaind ´ıa, L. Sebasti ´a, J. Gonz ´alez- Boticario, A. Rodr ´ıguez, S. Fern ´andez, J. D. Arias, and D. Borrajo, “Samap: An user-oriented adaptive system for planning tourist visits,” Expert Systems with Applications, vol. 34, no. 2, pp. 1318–1332, 2008

    2008

  3. [11]

    Research on travel route planing prob- lems based on greedy algorithm,

    G. Wang, Y . Wang, and W. Li, “Research on travel route planing prob- lems based on greedy algorithm,” in2024 4th International Conference on Electronic Information Engineering and Computer Science (EIECS). IEEE, 2024, pp. 755–758

    2024

  4. [12]

    Multi-objective optimal tourism route planning and design based on improved particle swarm optimization algorithm,

    J. Yan, “Multi-objective optimal tourism route planning and design based on improved particle swarm optimization algorithm,” inInternational Conference on Cognitive based Information Processing and Applica- tions. Springer, 2023, pp. 303–312

    2023

  5. [13]

    Application of ant colony optimization computing to a recommended travel itinerary planning system with repeatedly used nodes,

    S.-T. Chen, T.-H. Wu, R.-J. Ye, L.-C. Lee, W.-Y . Huang, Y .-H. Lin, and B.-Y . Wang, “Application of ant colony optimization computing to a recommended travel itinerary planning system with repeatedly used nodes,”Applied Sciences, vol. 13, no. 24, p. 13221, 2023

    2023

  6. [14]

    A systematic literature review for the tourist trip design problem: Extensions, solution techniques and future research lines,

    J. Ruiz-Meza and J. R. Montoya-Torres, “A systematic literature review for the tourist trip design problem: Extensions, solution techniques and future research lines,”Operations Research Perspectives, vol. 9, p. 100228, 2022

    2022

  7. [15]

    Constraint- objective cooperative coevolution for large-scale constrained optimiza- tion,

    P. Xu, W. Luo, X. Lin, J. Zhang, Y . Qiao, and X. Wang, “Constraint- objective cooperative coevolution for large-scale constrained optimiza- tion,”ACM Transactions on Evolutionary Learning and Optimization, vol. 1, no. 3, pp. 1–26, 2021

    2021

  8. [16]

    Designing benchmark problems for large-scale continuous optimization,

    M. N. Omidvar, X. Li, and K. Tang, “Designing benchmark problems for large-scale continuous optimization,”Information Sciences, vol. 316, pp. 419–436, 2015

    2015

  9. [17]

    Cooperative coevolution for cross-city itinerary planning,

    Z. Zhang, P. Xu, Z. Wang, and W. Luo, “Cooperative coevolution for cross-city itinerary planning,” inInternational Conference on Intelligent Information Processing. Springer, 2024, pp. 382–391

    2024

  10. [18]

    Generalized traveling salesman problem through n sets of nodes,

    S. Srivastava, S. Kumar, R. Garg, and P. Sen, “Generalized traveling salesman problem through n sets of nodes,”CORS journal, vol. 7, no. 2, p. 97, 1969

    1969

  11. [19]

    The clustered traveling salesman problem,

    J. A. Chisman, “The clustered traveling salesman problem,”Computers & Operations Research, vol. 2, no. 2, pp. 115–119, 1975

    1975

  12. [20]

    On planning sightseeing tours with tripbuilder,

    I. R. Brilhante, J. A. Macedo, F. M. Nardini, R. Perego, and C. Renso, “On planning sightseeing tours with tripbuilder,”Information Processing & Management, vol. 51, no. 2, pp. 1–15, 2015

    2015

  13. [21]

    Challenges in real world sightseeing tour optimization using meta-heuristics,

    G. Jean-Marc, “Challenges in real world sightseeing tour optimization using meta-heuristics,” inProceedings of the 6th WSEAS International Conference on Evolutionary Computing, 2005, pp. 233–238

    2005

  14. [22]

    A local search- based non-dominated sorting genetic algorithm for solving a multi- objective medical tourism trip design problem considering the attrac- tiveness of trips,

    M. H. Kolaee, S. M. J. M. Al-e, A. Jabbarzadehet al., “A local search- based non-dominated sorting genetic algorithm for solving a multi- objective medical tourism trip design problem considering the attrac- tiveness of trips,”Engineering Applications of Artificial Intelligenc...

    2023

  15. [23]

    Uncertain multi-objective programming model of tourist route considering tourist preference,

    Z. Huang, Y . Ning, and F. Liu, “Uncertain multi-objective programming model of tourist route considering tourist preference,”Journal of In- dustrial and Management Optimization, vol. 20, no. 8, pp. 2640–2651, 2024

    2024

  16. [24]

    A cooperative coevolutionary approach to function optimization,

    M. A. Potter and K. A. De Jong, “A cooperative coevolutionary approach to function optimization,” inInternational Conference on Parallel Prob- lem Solving from Nature. Springer, 1994, pp. 249–257

    1994

  17. [25]

    Multiobjective optimization and coevolution,

    S. G. Ficici, “Multiobjective optimization and coevolution,”Multiobjec- tive Problem Solving from Nature: From Concepts to Applications, pp. 31–52, 2008

    2008

  18. [26]

    A non-cooperative game for faster convergence in cooperative coevolution for multi-objective optimiza- tion,

    L. M. Antonio and C. A. C. Coello, “A non-cooperative game for faster convergence in cooperative coevolution for multi-objective optimiza- tion,” in2015 IEEE Congress on Evolutionary Computation (CEC). IEEE, 2015, pp. 109–116

    2015

  19. [27]

    A cooperative coevolutionary multiobjective algorithm using non-dominated sorting,

    A. W. Iorio and X. Li, “A cooperative coevolutionary multiobjective algorithm using non-dominated sorting,” inGenetic and Evolutionary Computation Conference. Springer, 2004, pp. 537–548

    2004

  20. [28]

    A multiobjective cooperative coevolutionary algorithm for hyperspectral sparse unmixing,

    M. Gong, H. Li, E. Luo, J. Liu, and J. Liu, “A multiobjective cooperative coevolutionary algorithm for hyperspectral sparse unmixing,”IEEE Transactions on Evolutionary Computation, vol. 21, no. 2, pp. 234–248, 2016

    2016

  21. [29]

    A review of population-based metaheuristics for large-scale black-box global optimization—Part I,

    M. N. Omidvar, X. Li, and X. Yao, “A review of population-based metaheuristics for large-scale black-box global optimization—Part I,” 13 IEEE Transactions on Evolutionary Computation, vol. 26, no. 5, pp. 802–822, 2021

    2021

  22. [30]

    A survey on cooperative co-evolutionary algorithms,

    X. Ma, X. Li, Q. Zhang, K. Tang, Z. Liang, W. Xie, and Z. Zhu, “A survey on cooperative co-evolutionary algorithms,”IEEE Transactions on Evolutionary Computation, vol. 23, no. 3, pp. 421–441, 2018

    2018

  23. [31]

    Coevolutionary multiobjective evo- lutionary algorithms: Survey of the state-of-the-art,

    L. M. Antonio and C. A. C. Coello, “Coevolutionary multiobjective evo- lutionary algorithms: Survey of the state-of-the-art,”IEEE Transactions on Evolutionary Computation, vol. 22, no. 6, pp. 851–865, 2017

    2017

  24. [32]

    A large-scale continuous optimization benchmark suite with versatile coupled heterogeneous modules,

    P. Xu, W. Luo, X. Lin, J. Zhang, and X. Wang, “A large-scale continuous optimization benchmark suite with versatile coupled heterogeneous modules,”Swarm and Evolutionary Computation, vol. 78, p. 101280, 2023

    2023

  25. [33]

    MLFS-CCDE: multi-objective large-scale feature selection by cooperative coevolutionary differential evolution,

    H. Li, F. He, Y . Chen, and Y . Pan, “MLFS-CCDE: multi-objective large-scale feature selection by cooperative coevolutionary differential evolution,”Memetic Computing, vol. 13, pp. 1–18, 2021

    2021

  26. [34]

    Cooperative co-evolution for large-scale multi-objective air traffic flow management,

    T. Guo, Y . Mei, K. Tang, and W. Du, “Cooperative co-evolution for large-scale multi-objective air traffic flow management,”IEEE Transac- tions on Evolutionary Computation, vol. 28, no. 6, 1644–1658

  27. [35]

    Smart use of computational resources based on contribution for cooperative co-evolutionary algo- rithms,

    M. N. Omidvar, X. Li, and X. Yao, “Smart use of computational resources based on contribution for cooperative co-evolutionary algo- rithms,” inProceedings of the 13th Annual Conference on Genetic and Evolutionary Computation, 2011, pp. 1115–1122

    2011

  28. [36]

    Many-modal opti- mization by difficulty-based cooperative co-evolution,

    W. Luo, Y . Qiao, X. Lin, P. Xu, and M. Preuss, “Many-modal opti- mization by difficulty-based cooperative co-evolution,” in2019 IEEE Symposium Series on Computational Intelligence (SSCI). IEEE, 2019, pp. 1907–1914

    2019

  29. [37]

    Difficulty and contribution-based cooperative coevolution for large-scale optimization,

    P. Xu, W. Luo, X. Lin, Y . Chang, and K. Tang, “Difficulty and contribution-based cooperative coevolution for large-scale optimization,” IEEE Transactions on Evolutionary Computation, vol. 27, no. 5, pp. 1355–1369, 2023

    2023

  30. [38]

    A novel multi-objective co-evolutionary algorithm based on decomposition approach,

    Z. Liang, X. Wang, Q. Lin, F. Chen, J. Chen, and Z. Ming, “A novel multi-objective co-evolutionary algorithm based on decomposition approach,”Applied Soft Computing, vol. 73, pp. 50–66, 2018

    2018

  31. [39]

    Au- tomatic planning of multiple itineraries: A niching genetic evolution approach,

    T. Huang, Y .-J. Gong, Y .-H. Zhang, Z.-H. Zhan, and J. Zhang, “Au- tomatic planning of multiple itineraries: A niching genetic evolution approach,”IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 10, pp. 4225–4240, 2019

    2019

  32. [40]

    A fast and elitist multiobjective genetic algorithm: NSGA-II,

    K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan, “A fast and elitist multiobjective genetic algorithm: NSGA-II,”IEEE Transactions on Evolutionary Computation, vol. 6, no. 2, pp. 182–197, 2002

    2002

  33. [41]

    Multi-objective route planning of museum guide based on an improved NSGA-II algorithm,

    Y . Xu, Q. Guo, A. Tan, L. Xu, Y . Tu, and S. Liu, “Multi-objective route planning of museum guide based on an improved NSGA-II algorithm,” inJournal of Physics: Conference Series, vol. 1828. IOP Publishing, 2021, p. 012051

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.